Feature map and image kernel visualisation on a model trained on cifar10 dataset. Feature map visualisation on 2 convolutional layers (with activation maps); 1st conv layer --kernel shape= (4 ,4) --kernels number = 32 --strides shape = (1 ,1)
2nd conv layer --kernel shape = (1 ,1 ) --kernels number = 16 --strides shape = (2 ,2 )
1st feature map shape = (29 ,29 ,32) 2nd feature map shape = (15 ,15, 16)